通过分阶段切换大模型小模型，实现视频生成又快又保质。
FlowBlending: Stage-Aware Multi-Model Sampling for Fast and High-Fidelity Video Generation
- 根据时间步重要性动态切换大、小模型
- 速度提升1.65倍，计算量减少57.35%
- 兼容现有加速技术，适合高效视频生成需求
本文发现模型容量在视频生成的不同时间步影响不同：早期和晚期关键，中间阶段影响小。为此提出FlowBlending，一种分阶段多模型采样策略，在容量敏感阶段使用大模型（LTX-Video 2B/13B、WAN 2.1 1.3B/14B），中间阶段使用小模型。通过速度差异分析确定阶段边界，有效识别容量敏感区域。实验显示，该方法在保持大模型视觉质量、时序连贯性和语义一致性的同时，推理速度最快提升1.65倍，计算量减少57.35%。且与现有采样加速技术兼容，可再提速最多2倍。项目页面见：https://jibin86.github.io/flowblending_project_page。
原文摘要 · Abstract (English)
In this work, we show that the impact of model capacity varies across timesteps: it is crucial for the early and late stages but largely negligible during the intermediate stage. Accordingly, we propose FlowBlending, a stage-aware multi-model sampling strategy that employs a large model and a small model at capacity-sensitive stages and intermediate stages, respectively. We further introduce simple criteria to choose stage boundaries and provide a velocity-divergence analysis as an effective proxy for identifying capacity-sensitive regions. Across LTX-Video (2B/13B) and WAN 2.1 (1.3B/14B), FlowBlending achieves up to 1.65x faster inference with 57.35% fewer FLOPs, while maintaining the visual fidelity, temporal coherence, and semantic alignment of the large models. FlowBlending is also compatible with existing sampling-acceleration techniques, enabling up to 2x additional speedup. Project page is available at: https://jibin86.github.io/flowblending_project_page.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。